What is done with the folding results?

[Ericson_Mar]

Just wondering what is done after proteins are folded? Are these verified experimentally in “assays” etc. to see if worked right? Is it at a stage where the computer simulations are still being tested (and then modified)? Or are they actually using the results to conduct experiments with real proteins? On chips, with robots, in cells? Etc.? Or are they waiting for all the results to perform some "metabolic" chain experiments or something?

[uncle fuzzy]

Click the "Home" link at the top of this page, then the "Results" link at the top of that page. Good stuff.

[PurePunjabi]

Click the "Home" link at the top of this page, then the "Results" link at the top of that page. Good stuff.

Could you not just summarize it? am not reading all of that unless someones going to pay me.

[uncle_fungus]

Here's the layman's summary of paper 50:

When we study protein folding, we usually simulate just a single protein in a large box of water. This is similar to how experimentalists study protein folding; they use concentrations that are low enough that you can assume the proteins are not interacting with each other (i.e. sticking to each other or as I like to say “seeing each other”). This is a useful way to try to understand how proteins fold, but this is different from how things actually are inside a cell.

Inside a cell, the concentration of protein is very high (we call this macromolecular crowding). This crowding makes processes such as protein aggregation (unfolded proteins sticking together, i.e. what happens in Alzheimer’s disease) more likely. Additionally, cells suffer a number of types of environmental stress that may make it harder for proteins to stay folded. Because it can be much harder to fold in a cell than in a test tube, cells come equipped with special helper molecules called chaperonins. Chaperonins are like large cages that swallow unfolded (or misfolded) proteins and help them to fold correctly. This is a tricky issue because it is not known how chaperonins actually fold proteins. One idea is that by confining swallowed proteins, they essentially compress them into the correct (native) shape (i.e. confinement restricts the protein from being in more extended shapes, this
entropically stabilizes the native state). Another idea is that chaperonins somehow unfold proteins. You may be saying, “Wait a minute here, I thought chaperonins helped proteins to fold, not unfold!” The answer to this is that sometimes when a protein misfolds it gets trapped in the misfolded state.

These misfolded proteins can’t do their jobs properly and also tend to aggregate. So some say that a chaperonin swallows a misfolded protein and unfolds it. This effectively gives it another chance to fold to the correct shape. Other researchers in the past have done simulations where they show that if you confine a protein inside a small volume, it is more likely to fold. This supports the first idea I mentioned above. These simulations however, do not really take into account the fact that the protein is surrounded by hundreds of water molecules and these water molecules may have an impact on folding. Not too long ago Eric Sorin in our group published a paper showing that if you confine a very simple protein (an isolated alpha helix) inside a
carbon nanotube (a long hollow tube made of inert carbon) and explicitly include the water molecules, the protein unfolds. This agrees with the second idea I mentioned above.

So our goal was to take one step towards sorting all of this out. We wanted to have two sets of simulations that we can directly compare to each other to see how these two effects contribute to overall folding behavior. We took a small protein (one that we have watched fold in previous work) and put it inside a continuous repulsive bubble (a nanopore). The bubble is designed so that if something gets too close, it gently pushes it away. I included the water explicitly in the simulations. Now the trick for this work is that we can choose what the bubble can see (interact with) and what it can’t. For one set of simulations the bubble can see the protein but not the water. This means that the protein is confined, but the water can pass freely through the bubble as if it wasn’t there. We then did another set of simulations where both the protein and the water both see the bubble (i.e. nothing can pass in or out of the bubble). We then used techniques invented in the group to compare these two sets of simulations to each other. If you look at figure 1 in the paper there is a nice little image of this setup.

What we found is that when only protein is confined (the bubble only sees the protein), the protein is more likely to fold. But, when both protein and water are the protein is more likely to unfold. We also found that when the protein unfolds, it collapses to a small globule that is similar in size to the native state, but different in shape (specifically, it looses most of its secondary structure). This is different from what happens when a protein unfolds in a regular simulation. Finally, we redo this for different sized bubbles to see how the size of the bubble affects this. We found that for very, very small bubbles confinement helps the protein to fold. But for moderate to larger bubbles, confinement causes the protein to unfold.

Using these findings, we postulate that chaperonins may use both mechanisms (squishing the protein to its correct shape, or unfolding it to give it another chance at folding) to help proteins fold, but they may favor one mechanism or the other depending on the size of the protein. Of course the model we use is very simple and neglects some things that may be very important for chaperonin mechanism. For this reason Jeremy England (another member of our group) and I are setting up some newer simulations where the bubble is more like an actual chaperonin. We are also working on actual simulations of a protein inside a real chaperonin (not a simple bubble). These simulations are especially challenging because the system size is
huge and the timescales are very long (even for folding@home).

Del Lucent

[ppetrone]

Hi Ericson_Mar,

The data that come from F@H simulations is processed and analyzed. We withdraw patterns that translate as results that pile up as papers that get published and add up to what we know about chemistry, biology and biophysics. These results come generally in the form of predictions that motivate new experiments because we are "looking" at things that happen at a resolution and timescales off the range of many experiments. New experiments inspire new simulations, and therefore more data and the wheel keep going forward.

Some simulations like the projects in which I am working on (Ribosome related), do not study actual folding mechanisms but interactions between small molecules and important cellular machines like the ribosome or the translocon channel. Other people in the lab simulate membranes (like Peter) for example, or Del who studies chaperones.

In any case, the folding results produce knowledge.

Thank you for contributing to the project!

paula